Mapping of Dwellings in IDP/Refugee Settlements from Very High-Resolution Satellite Imagery Using a Mask Region-Based Convolutional Neural Network

Using spatial video and deep learning for automated mapping of ground-level context in relief camps

BACKGROUND

The creation of relief camps following a disaster, conflict or other form of externality often generates additional health problems. The density of people in a highly stressed environment with questionable safe food and water access presents the potential for infectious disease outbreaks. These camps are also not static data events but rather fluctuate in size, composition, and level and quality of service provision. While contextualized geospatial data collection and mapping are vital for understanding the nature of these camps, various challenges, including a lack of data at the required spatial or temporal granularity, as well as the issue of sustainability, can act as major impediments. Here, we present the first steps toward a deep learning-based solution for dynamic mapping using spatial video (SV).

METHODS

We trained a convolutional neural network (CNN) model on a SV dataset collected from Goma, Democratic Republic of Congo (DRC) to identify relief camps from video imagery. We developed a spatial filtering approach to tackle the challenges associated with spatially tagging objects such as the accuracy of global positioning system and positioning of camera. The spatial filtering approach generates smooth surfaces of detection, which can further be used to capture changes in microenvironments by applying techniques such as raster math.

RESULTS

The initial results suggest that our model can detect temporary physical dwellings from SV imagery with a high level of precision, recall, and object localization. The spatial filtering approach helps to identify areas with higher concentrations of camps and the web-based tool helps to explore these areas. The longitudinal analysis based on applying raster math on the detection surfaces revealed locations, which had a considerable change in the distribution of tents over space and time.

CONCLUSIONS

The results lay the groundwork for automated mapping of spatial features from imagery data. We anticipate that this work is the building block for a future combination of SV, object identification and automatic mapping that could provide sustainable data generation possibilities for challenging environments such as relief camps or other informal settlements.

Using machine learning and remote sensing to track land use/land cover changes due to armed conflict

Armed conflicts have detrimental impacts on the environment, including land systems. The prevailing understanding of the relation between Land Use/Land Cover (LULC) and armed conflict fails to fully recognize the complexity of their dynamics - a shortcoming that could undermine food security and sustainable land/water resources management in conflict settings. The Syrian portion of the transboundary Orontes River Basin (ORB) has been a site of violent conflict since 2013. Correspondingly, the Lebanese and Turkish portions of the ORB have seen large influxes of refugees. A major challenge in any geoscientific investigation in this region, specifically the Syrian portion, is the unavailability of directly-measured "ground truth" data. To circumvent this problem, we develop a novel methodology that combines remote sensing products, machine learning techniques and quasi-experimental statistical analysis to better understand LULC changes in the ORB between 2004 and 2022. Through analysis of the resulting annual LULC maps, we can draw several quantitative conclusions. Cropland areas decreased by 21-24 % in Syria's conflict hotspot zones after 2013, whereas a 3.4-fold increase was detected in Lebanon. The development of refugee settlements was also tracked in Lebanon and on the Syrian/Turkish borders, revealing different LULC patterns that depend on settlement dynamics. The results highlight the importance of understanding the heterogenous spatio-temporal LULC changes in conflict-affected and refugee-hosting countries. The developed methodology is a flexible, cloud-based approach that can be applied to wide variety of LULC investigations related to conflict, policy and climate.